© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Initial Gemma-2B-IT SAEs
    5. Gemma-2B-IT
    7. Residual Stream
    9. 12-RES-JB
    11. 5034
    Prev
    Next
    INDEX
    Explanations

    references to violence or aggression, particularly related to teaching or lessons

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    HuggingFaceFW/fineweb
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.12.hook_resid_post
    Hook Layer
    12
    Architecture
    standard
    Context Size
    1,024
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    604
    +0.10
    0.3%
    1531
    +0.08
    0.2%
    1842
    +0.07
    0.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    392
    +0.10
    0.03
    940
    +0.08
    0.02
    1114
    +0.07
    0.03
    Negative Logits
     increa
    -1.68
     depic
    -1.66
     fta
    -1.64
     purcha
    -1.64
     milf
    -1.63
     strick
    -1.61
     guarante
    -1.60
     snoopy
    -1.59
     ?...
    -1.59
     attemp
    -1.58
    POSITIVE LOGITS
    twimg
    0.65
     teach
    0.57
    Xar
    0.57
    Muerte
    0.57
    belongs
    0.56
    Quien
    0.55
    llary
    0.55
     inaugu
    0.54
    GIVEREF
    0.54
    Élet
    0.54
    Activations Density 0.396%

    No Known Activations